Handbook of Monetary Economics - Read Online

Book Preview

Handbook of Monetary Economics

You've reached the end of this preview. Sign up to read more!
Page 1 of 1

Handbook of Monetary Economics

First Edition

Benjamin M. Friedman

Michael Woodford

VOLUME 3B

Amsterdam • Boston • Heidelberg • London • New York • Oxford

Paris • San Diego • San Francisco • Singapore • Sydney • Tokyo

North-Holland is an imprint of Elsevier

Table of Contents

Cover image

Title page

Copyright page

Contributors

Preface

Part Four: Optimal Monetary Policy

Chapter 13: The Optimal Rate of Inflation

Abstract

1 INTRODUCTION

2 MONEY DEMAND AND THE OPTIMAL RATE OF INFLATION

3 MONEY DEMAND, FISCAL POLICY AND THE OPTIMAL RATE OF INFLATION

4 FAILURE OF THE FRIEDMAN RULE DUE TO UNTAXED INCOME: THREE EXAMPLES

5 A FOREIGN DEMAND FOR DOMESTIC CURRENCY AND THE OPTIMAL RATE OF INFLATION

6 STICKY PRICES AND THE OPTIMAL RATE OF INFLATION

7 THE FRIEDMAN RULE VERSUS PRICE-STABILITY TRADE-OFF

8 DOES THE ZERO BOUND PROVIDE A RATIONALE FOR POSITIVE INFLATION TARGETS?

9 DOWNWARD NOMINAL RIGIDITY

10 QUALITY BIAS AND THE OPTIMAL RATE OF INFLATION

11 CONCLUSION

APPENDIX

Chapter 14: Optimal Monetary Stabilization Policy

Abstract

1 INTRODUCTION

2 OPTIMAL POLICY IN A CANONICAL NEW KEYNESIAN MODEL

3 STABILIZATION AND WELFARE

4 GENERALIZATIONS OF THE BASIC MODEL

5 RESEARCH AGENDA

Chapter 15: Simple and Robust Rules for Monetary Policy

Abstract

1 INTRODUCTION

2 HISTORICAL BACKGROUND

3 USING MODELS TO EVALUATE SIMPLE POLICY RULES

4 ROBUSTNESS OF POLICY RULES

5 OPTIMAL POLICY VERSUS SIMPLE RULES

6 LEARNING FROM EXPERIENCE BEFORE, DURING AND AFTER THE GREAT MODERATION

7 CONCLUSION

Chapter 16: Optimal Monetary Policy in Open Economies

Abstract

1 INTRODUCTION AND OVERVIEW

PART I: OPTIMAL STABILIZATION POLICY AND INTERNATIONAL RELATIVE PRICES WITH FRICTIONLESS ASSET MARKETS

2 A BASELINE MONETARY MODEL OF MACROECONOMIC INTERDEPENDENCE

3 THE CLASSICAL VIEW: DIVINE COINCIDENCE IN OPEN ECONOMIES

4 SKEPTICISM ON THE CLASSICAL VIEW: LOCAL CURRENCY PRICE STABILITY OF IMPORTS

5 DEVIATIONS FROM POLICY COOPERATION AND CONCERNS WITH COMPETITIVE DEVALUATIONS

PART II: CURRENCY MISALIGNMENTS AND CROSS-COUNTRY DEMAND IMBALANCES

6 MACROECONOMIC INTERDEPENDENCE UNDER ASSET MARKET IMPERFECTIONS

7 CONCLUSIONS

Part Five: Constraints on Monetary Policy

Chapter 17: The Interaction Between Monetary and Fiscal Policy

Abstract

1 INTRODUCTION

2 POSITIVE THEORY OF PRICE STABILITY

3 NORMATIVE THEORY OF PRICE STABILITY: IS PRICE STABILITY OPTIMAL?

Chapter 18: The Politics of Monetary Policy

Abstract

1 INTRODUCTION

2 RULES VERSUS DISCRETION

3 CENTRAL BANK INDEPENDENCE

4 POLITICAL BUSINESS CYCLES

5 CURRENCY UNIONS

6 THE EURO

7 CONCLUSION

APPENDIX

Chapter 19: Inflation Expectations, Adaptive Learning and Optimal Monetary Policy

Abstract

1 INTRODUCTION

2 RECENT DEVELOPMENTS IN PRIVATE-SECTOR INFLATION EXPECTATIONS

3 A SIMPLE NEW KEYNESIAN MODEL OF INFLATION DYNAMICS UNDER RATIONAL EXPECTATIONS

4 MONETARY POLICY RULES AND STABILITY UNDER ADAPTIVE LEARNING

5 OPTIMAL MONETARY POLICY UNDER ADAPTIVE LEARNING

6 SOME FURTHER REFLECTIONS

7 CONCLUSIONS

Chapter 20: Wanting Robustness in Macroeconomics

Abstract

1 INTRODUCTION

2 KNIGHT, SAVAGE, ELLSBERG, GILBOA-SCHMEIDLER, AND FRIEDMAN

3 FORMALIZING A TASTE FOR ROBUSTNESS

4 CALIBRATING A TASTE FOR ROBUSTNESS

5 LEARNING

6 ROBUSTNESS IN ACTION

7 CONCLUDING REMARKS

APPENDIX

Part Six: Monetary Policy in Practice

Chapter 21: Monetary Policy Regimes and Economic Performance: The Historical Record, 1979–2008

Abstract

1 INTRODUCTION

2 MONETARY TARGETRY, 1979–1982

3 INFLATION TARGETS

4 THE NICE YEARS, 1993–2006

5 EUROPE AND THE TRANSITION TO THE EURO

6 JAPAN

7 FINANCIAL STABILITY AND MONETARY POLICY DURING THE FINANCIAL CRISIS

8 CONCLUSIONS AND IMPLICATIONS FOR FUTURE CENTRAL BANK POLICIES

APPENDIX

Chapter 22: Inflation Targeting

Abstract

1 INTRODUCTION

2 HISTORY AND MACROECONOMIC EFFECTS

3 THEORY

4 PRACTICE

5 FUTURE

Chapter 23: The Performance of Alternative Monetary Regimes

Abstract

1 INTRODUCTION

2 SOME SIMPLE EVIDENCE

3 PREVIOUS WORK ON INFLATION TARGETING

4 THE EURO

5 THE ROLE OF MONETARY AGGREGATES

6 HARD CURRENCY PEGS

7 CONCLUSION

APPENDIX

Chapter 24: Implementation of Monetary Policy: How Do Central Banks Set Interest Rates?

Abstract

1 INTRODUCTION

2 FUNDAMENTAL ISSUES IN THE MODE OF WICKSELL

3 THE TRADITIONAL UNDERSTANDING OF HOW THEY DO THAT

4 OBSERVED RELATIONSHIPS BETWEEN RESERVES AND THE POLICY INTEREST RATE

5 HOW, THEN, DO CENTRAL BANKS SET INTEREST RATES?

6 EMPIRICAL EVIDENCE ON RESERVE DEMAND AND SUPPLY WITHIN THE MAINTENANCE PERIOD

7 NEW POSSIBILITIES FOLLOWING THE 2007–2009 CRISIS

8 CONCLUSION

Chapter 25: Monetary Policy in Emerging Markets

Abstract

1 INTRODUCTION

2 WHY DO WE NEED DIFFERENT MODELS FOR EMERGING MARKETS?

3 GOODS MARKETS, PRICING, AND DEVALUATION

4 INFLATION

5 NOMINAL TARGETS FOR MONETARY POLICY

6 EXCHANGE RATE REGIMES

7 PROCYCLICALITY

8 CAPITAL FLOWS

9 Crises in emerging markets

10 SUMMARY OF CONCLUSIONS

Index-Volume 3B

Index-Volume 3A

Copyright

North-Holland in an imprint of Elsevier

525 B Street, Suite 1800, San Diego, CA 92101-4495, USA

Radarweg 29, 1000 AE Amsterdam, The Netherlands

First edition 2011

Copyright © 2011 Elsevier B.V. All rights reserved

No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means electronic, mechanical, photocopying, recording or otherwise without the prior written permission of the publisher

Permissions may be sought directly from Elsevier’s Science & Technology Rights Department in Oxford, UK: phone (+ 44) (0) 1865 843830; fax (+ 44) (0) 1865 853333; email: permissions@elsevier.com. Alternatively you can submit your request online by visiting the Elsevier web site at http://elsevier.com/locate/permissions, and selecting Obtaining permission to use Elsevier material

Notice

No responsibility is assumed by the publisher for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions or ideas contained in the material herein. Because of rapid advances in the medical sciences, in particular, independent verification of diagnoses and drug dosages should be made

Library of Congress Cataloging-in-Publication Data

A catalog record for this book is available from the Library of Congress

British Library Cataloguing in Publication Data

A catalogue record for this book is available from the British Library

ISBN Vol 3B: 978-0-444-53454-5

ISBN Vol 3A: 978-0-444-53238-1

SET ISBN: 978-0-444-53470-5

For information on all North-Holland publications

visit our website at elsevierdirect.com

Printed and bound in the USA

11 12 13 10 9 8 7 6 5 4 3 2 1

Contributors

Alberto Alesina     Harvard University

Laurence Ball     Johns Hopkins University

Luca Benati     European Central Bank

Matthew Canzoneri     Georgetown University

Giancarlo Corsetti     Cambridge University

Robert Cumby     Georgetown University

Luca Dedola     European Central Bank

Behzad Diba     Georgetown University

Jeffrey Frankel     Harvard University

Benjamin M. Friedman     Harvard University

Vitor Gaspar     Banco de Portugal

Charles Goodhart     London School of Economics, Financial Markets Group

Lars Peter Hansen     University of Chicago

Kenneth N. Kuttner     Williams College

Sylvain Leduc     Federal Reserve Bank of San Francisco

Thomas J. Sargent     New York University

Stephanie Schmitt-Grohé     Columbia University

Frank Smets     European Central Bank

Andrea Stella     Harvard University

Lars E.O. Svensson     Sveriges Riksbank

John B. Taylor     Stanford University

Martín Uribe     Columbia University

David Vestin     Sveriges Riksbank

John C. Williams     Federal Reserve Bank of San Francisco

Michael Woodford     Columbia University

Preface

These new volumes supplement and bring up to date the original Handbook of Monetary Economics (Volumes I and II of this series), edited by Benjamin Friedman with Frank Hahn. It is now twenty years since the publication of those earlier volumes, so a reconsideration of the field is timely if not overdue. Some of the topics covered in the previous volumes of Handbook of Monetary Economics were updated in the Handbook of Macroeconomics, edited by Michael Woodford with John Taylor, but it is now ten years since the publication of those volumes as well. Further, that publication, with its broader focus on macroeconomics, could not fully substitute for a new edition of the Handbook of Monetary Economics. The subject here is macroeconomics, to be sure, but it is monetary macroeconomics.

Publication of a handbook in some area of intellectual inquiry usually means that researchers in the field have made substantial progress that is worth not only reviewing but also adding, in summary form, to the canonical presentation of work made conveniently available to students and other interested scholars. As the 25 chapters included in these new volumes make clear, this has certainly been the case in monetary macroeconomics. While many chapters of both the 1990 Handbook of Monetary Economics and the 2000 Handbook of Macroeconomics will remain valuable resources, the pace of recent progress has been such that a summary from even as recently as a decade ago is incomplete in many important respects. These new volumes are intended to fill that gap.

Publication of a handbook also often means that a field has reached a sufficient stage of maturity so that it is safe to take stock without concern that new ideas, or the press of external events, will soon result in significant new directions. Today, however, the opposite is likely to be true in monetary macroeconomics. The extraordinary economic and financial events of 2007–2010 seem highly likely to prod researchers to consider new lines of thinking, and to evaluate old ones against new bodies of evidence that in many key respects differ sharply from prior experience. It is obviously too early for us to anticipate what the full consequences of such reconsideration would be. We believe, however, that it is valuable to take stock of the state of the field before the deluge. Further, a number of the chapters included here present early attempts to pursue lines of inquiry suggested by the 2007–2010 experience.

Developments in the world economy since the publication of the earlier volumes of this Handbook provided much new ground for economic thinking, even prior to the recent crisis, and these had already spurred significant developments in monetary macroeconomics as well. Among the notable monetary experiments of the past two decades, we should mention two in particular. The creation of a monetary union in Europe has not only introduced a new major world currency and a new central bank, but has revived interest in the theory of monetary unions and optimal currency areas and raised novel questions about the degree to which it is possible to separate monetary policy from fiscal policy and from financial supervision (the latter issues are handled at a completely different level of government in the Euro Zone). And the spread of inflation targeting as an approach to the conduct of monetary policy — first adopted mainly by members of the OECD, now increasingly popular among emerging market economies as well, but still resisted by a number of highly visible central banks (including, most clearly, the U.S. Federal Reserve System) — has brought not only a stronger degree of emphasis on inflation stabilization as a policy goal but also greater explicitness about central banks’ policy targets and a more integrated role for quantitative modeling in policy deliberations. It has also changed central banks’ communications with the public about those deliberations. Both of these developments have been the subject of extensive scholarly analysis, both theoretical and empirical, and they are treated in detail in several chapters of these new volumes.

The past two decades have witnessed important methodological advances in monetary macroeconomics as well. One of the more notable of these has been the development of empirical dynamic stochastic general equilibrium (DSGE) models that incorporate serious (although also seriously incomplete) efforts to capture the monetary policy transmission mechanism. While these models are doubtless still at a fairly early stage of development, and the adequacy of current-generation DSGE models for practical policy analysis remains a topic of lively debate, for at least the past decade they have been an important focus of research efforts, particularly in central banks around the world and in other policy institutions. Quite a few of the chapters included here rely on these models, while several others examine these models’ structure and the methods used to estimate and evaluate them, with particular emphasis on the account that they give of the transmission mechanism for monetary policy.

There have also been important changes in the methods used to assess the empirical realism of particular models. One important development has been the increasing use of structural vector autoregression methodology to estimate the effects of monetary policy shocks under relatively weak theoretical assumptions. The chapter on this topic in the Handbook of Macroeconomics (Chapter 7; Christiano, Eichenbaum, and Evans, 1999) provides a sufficient exposition of this method; but several of the chapters included in these volumes illustrate how this method is now routinely used in applied work. Another notable development in empirical methodology has been increasing use by macroeconomists of individual or firm-level data sets, and not simply aggregate time series, as sources of evidence about aspects of behavior that are central to macroeconomic models. Some of the work surveyed in these new volumes illustrates this importation of micro-level data into monetary macroeconomics.

Finally, there have been important methodological innovations in monetary policy analysis as well. Research on monetary policy rules has exploded over this period, having received considerable impetus from the celebrated proposal of the Taylor rule (Taylor, 1993), which not only suggested the possibility that some fairly simple rules might have desirable properties, but also indicated that some aspects of the behavior of actual central banks might be usefully characterized in terms of simple rules. Among other notable developments, an active literature over the past decade has assessed proposed rules for the conduct of monetary policy in terms of their implications for welfare as measured by the private objectives (household utility) that underlie the behavioral relations in microfounded models of the monetary transmission mechanism — essentially applying to monetary policy the method that had already become standard in the theory of public finance. Many of the chapters in these new Handbook volumes address these issues, and others related to them as well.

The events of the years immediately preceding publication of these new Handbook volumes have presented further challenges and opportunities for research in much of economics, but in monetary macroeconomics in particular. The 2007–2010 financial crisis and economic downturn constituted one of the most significant sequences of economic dislocations since World War II. In many countries the real economic costs — costs in terms of reduced production, lost jobs, shrunken investment, and foregone incomes and profits — exceeded those of any prior post-war decline. It was in the financial sector, however, that this latest episode primarily stood out. The collapse of major financial firms, the decline in asset values and consequent destruction of paper wealth, the interruption of credit flows, the loss of confidence both in firms and in credit market instruments, the fear of default by counterparties, and above all the intervention by central banks and other governmental institutions, were extraordinary.

Large-scale and unusual events often present occasions for introspection and learning, especially when they bring unwanted consequences. David Hume (1987), residing in Edinburgh during the Scottish banking crisis of 1772, wrote of that distressing sequence of events to his close friend Adam Smith. After recounting the bank failures, spreading unemployment, and Suspicion surrounding yet other industrial firms as well as banks, including even the Bank of England, Hume asked his friend, Do these Events any-wise affect your Theory? They certainly did. In The Wealth of Nations, published just four years later, Smith took the 1772 crisis into account in describing the interrelation of banking and nonfinancial economic activity and recommended a set of policy interventions that he thought would preclude or at least soften such disastrous episodes in the future.

The field of monetary macroeconomics has always been especially subject to just this kind of influence stemming from events in the world of which researchers are attempting to gain an understanding. Even the very origins of the field reflect the influence of real-world events. For all practical purposes it was the depression of the 1930s that created monetary macroeconomics as a recognizable component within the broader discipline, placing the obvious fact of limited price flexibility, and its consequences, at the center of the field’s attention, and introducing new intellectual constructs like aggregate demand. In the 1970s, as high inflation rates became both widespread and chronic across most industrialized economies, further new constructs such as dynamic inconsistency, again together with its consequences, profoundly influenced the field’s approach to issues of monetary policy. In the 1980s, the experience of disinflation led the field to change its direction and focus once again, as the costs associated with disinflation in many countries contradicted key lines of thinking spawned during the prior decade, and it was difficult to identify first-order differences in the disinflation experiences of countries that had pursued different policy paths and under different policy institutions.

There is no reason to expect the events of 2007–2010 to have any lesser impact. One influence that is already evident in new work in the field, and reflected in several of the chapters included in these new Handbook volumes, is an enhanced focus on credit; that is, the liability side of the balance sheets of households and firms and, conversely, the asset side (as opposed to the deposit, or money side) of the balance sheets of banks and other financial institutions. The reason is plain enough. In most economies that experienced severe crises and economic downturns in 2007–2010, the quantity of money did not decline and there was no evident scarcity of reserves supplied to the banking system by the central bank. Instead, what mattered, both in the origins of the crisis and for its consequences for nonfinancial economic activity, was the volume and price and availability of credit.

Another aspect of the crisis that has inspired new lines of research, also reflected in some of the chapters included in these new volumes, is the role of nonbank financial institutions. Traditional monetary economics, with its emphasis on the presumed central role of households’ and firms’ holdings of deposits as assets, naturally focused on deposit-issuing institutions. In some economies in recent decades, nonbank institutions began to issue deposit-like instruments, and therefore they too became of interest; but the volumes involved were normally small, and as an intellectual matter it was easy enough to consider these firms merely as a different form of bank. By contrast, once the emphasis shifts to the credit side of financial activity, the path is open for entertaining a key role for institutions that are very unlike banks and that may issue no deposit-like liabilities at all. At the same time, it becomes all the more important to understand the role played by prevailing institutions, including matters of financial regulation as well as more general aspects of business organization and practice (limited liability and the consequent distortion of incentives, broadly dispersed stockownership and the consequent principal-agent conflicts, and the like). Several of the chapters included here summarize the most recent research, or present entirely new research, along just these lines.

Yet further lines of inquiry motivated by the 2007–2010 experience remain sufficiently new, or as yet untried in a satisfactorily fleshed-out way, or even fundamentally uncertain, that it is still too early for these new Handbook volumes to reflect them. Will the experience of pricing of some credit market instruments — most obviously, claims against U.S. residential mortgages, but many others besides — lead to a broader questioning of what have until now been standard presumptions about rationality of asset markets? Will new theoretical advances make it possible to render the degree of market rationality, in this and other contexts, endogenous with respect to either economic outcomes or economic policy arrangements? Will the surprising (to many economists) use of discretionary anti-cyclical fiscal policy in many countries, or the sharp and seemingly sudden deterioration in governments’ fiscal positions, lead to renewed interest in fiscal-monetary connections, possibly with new normative implications? Most generally of all, will the experience of the deepest and longest lasting economic downturn in six decades lead to new thinking about the business cycle itself, including its origins as well as potential policy remediation?

As of 2010, the answer in each case is that no one knows. All that seems certain, given past experience, is that monetary macroeconomics will continue to evolve — and, we trust, to progress. In another decade, or two, there will be room for yet a further Handbook to supplement these new volumes. But for now, the 25 chapters published for the first time here speak to the status of a field that has been and will continue to be central to the discipline of economics. We hope students of the field, both new and experienced, will learn from them.

Our foremost debt in presenting these new Handbook volumes is to the authors who have contributed their work to be published here. Their own research and their review of the research of others is ample testimony to the effort they have put into this project, and we are grateful to every one of them for it.

We are also grateful to many others who have also added their efforts to this endeavor. Each of the chapters published here was presented, in early draft form, at one of two conferences held in the fall of 2009: one hosted by the Board of Governors of the Federal Reserve System and the other by the European Central Bank. We thank the Board and the ECB for their support of this project and for their generous hospitality. We are also grateful to the economists at these two institutions who took the lead in organizing these two events: at the Federal Reserve Board, Christopher Erceg, Michael Kiley, and Andrew Levin; and at the ECB, Frank Smets and Oreste Tristani. The planning of these conferences required an enormous amount of personal effort on their part, and we certainly appreciate it. We also thank Sue Williams at the Federal Reserve Board and Iris Bettenhauser at the ECB for the efficient and friendly staff support that they rendered.

The presentation of each draft chapter, at one or the other of these two conferences, involved a prepared response by a designated discussant. We are especially grateful to the over two dozen fellow economists who devoted their efforts to offering extremely thoughtful discussions that in most cases turned out to be both highly constructive and helpful. Their commentaries are not explicitly included in these volumes, but the ideas that they suggested are well reflected in the revised chapters published here. With few exceptions, these chapters are better — better thought out, better organized, better written, and more comprehensive in surveying the relevant research in their assigned areas — because of the comments that the authors received at the conferences.

Finally, we are grateful to Kenneth Arrow and Michael Intriligator, the long-time general editors of this Handbook series, for urging us to undertake these new volumes of the Handbook of Monetary Economics. We would not have done so without their encouragement.

Benjamin M. Friedman, Harvard University

Michael Woodford, Columbia University, May, 2010

REFERENCES

Christiano LJ, Eichenbaum M, Evans CL. Monetary policy shocks: What have we learned and to what end?. In: Amsterdam: Elsevier; . In: Taylor JB, Woodford M, eds. Handbook of macroeconomics. 1999;vol. 1A.

Hume D. Letter to Adam Smith, 3 September 1772. In: Mossner EC, Ross IS, eds. Correspondence of Adam Smith. Oxford, UK: Oxford University Press; 1987:131.

Taylor JB. Discretion versus policy rules in practice. Carnegie-Rochester Conference Series in Public Policy. 1993;39:195–214.

Part Four

Optimal Monetary Policy

Chapter 13

The Optimal Rate of Inflation

Stephanie Schmitt-Grohé*    * Columbia University, CEPR, and NBER

Martín Uribe**    ** Columbia University and NBER

Abstract

Observed inflation targets around the industrial world are concentrated at two percent per year. This chapter investigates the extent to which the observed magnitudes of inflation targets are consistent with the optimal rate of inflation predicted by leading theories of monetary non-neutrality. We find that consistently those theories imply that the optimal rate of inflation ranges from minus the real rate of interest to numbers insignificantly above zero. Furthermore, we argue that the zero bound on nominal interest rates does not represent an impediment for setting inflation targets near or below zero. Finally, we find that central banks should adjust their inflation targets upward by the size of the quality bias in measured inflation only if hedonic prices are more sticky than nonquality-adjusted prices.

JEL classification: E31, E4, E5

Keywords

Downward Nominal Rigidities

Foreign Demand for Money

Friedman Rule

Quality Bias

Ramsey Policy

Sticky-Prices

Zero Bound

Contents

1. Introduction   654

2. Money Demand and The Optimal Rate of Inflation   658

2.1 Optimality of the friedman rule with lump-sum taxation   662

3. Money Demand, Fiscal Policy and The Optimal Rate of Inflation   664

3.1 The primal form of the competitive equilibrium   665

3.2 Optimality of the Friedman rule with distortionary taxation   666

4. Failure of The Friedman Rule Due To Untaxed Income: Three Examples   667

4.1 Decreasing returns to scale   668

4.2 Imperfect competition   670

4.3 Tax evasion   672

5. A Foreign Demand For Domestic Currency and The Optimal Rate of Inflation   675

5.1 The model   675

5.2 Failure of the Friedman rule   677

5.3 Quantifying the optimal deviation from the Friedman rule   679

5.4 Lump-sum taxation   681

6. Sticky Prices and The Optimal Rate of Inflation   684

6.1 A sticky-price model with capital accumulation   684

6.2 Optimality of zero inflation with production subsidies   689

6.3 Optimality of zero inflation without production subsidies   690

6.4 Indexation   693

7. The Friedman Rule Versus Price-Stability Trade-Off   695

7.1 Sensitivity of the optimal rate of inflation to the degree of price stickiness   697

7.2 Sensitivity of the optimal rate of inflation to the size and elasticity of money demand   700

8. Does The Zero Bound Provide A Rationale For Positive Inflation Targets?   701

9. Downward Nominal Rigidity   704

10. Quality Bias and The Optimal Rate of Inflation   706

10.1 A simple model of quality bias   707

10.2 Stickiness in nonquality-adjusted prices   709

10.3 Stickiness in quality-adjusted prices   713

11. Conclusion   715

References   720

1 INTRODUCTION

The inflation objectives of virtually all central banks around the world are significantly above zero. Among monetary authorities in industrial countries that self-classify as inflation targeters, for example, inflation targets are concentrated at a level of 2% per year (Table 1). Inflation objectives are about one percentage point higher in inflation-targeting emerging countries. The central goal of this chapter is to investigate the extent to which the observed magnitudes of inflation targets are consistent with the optimal rate of inflation predicted by leading theories of monetary non-neutrality. We find that consistently those theories imply that the optimal rate of inflation ranges from minus the real rate of interest to numbers insignificantly above zero. Our findings suggest that the empirical regularity regarding the size of inflation targets cannot be reconciled with the optimal long-run inflation rates predicted by existing theories. In this sense, the observed inflation objectives of central banks pose a puzzle for monetary theory.

Table 1

Inflation Targets Around the World

Source: World Economic Outlook, 2005 Table 4.1.

In the existing literature, two major sources of monetary non-neutrality govern the determination of the optimal long-run rate of inflation. One source is a nominal friction stemming from a demand for fiat money. The second source is given by the assumption of price stickiness.

In monetary models in which the only nominal friction takes the form of a demand for fiat money for transaction purposes, optimal monetary policy calls for minimizing the opportunity cost of holding money by setting the nominal interest rate to zero. This policy, also known as the Friedman rule, implies an optimal rate of inflation that is negative and equal in absolute value to the real rate of interest. If the long-run real rate of interest lies, say, between 2 and 4%, the optimal rate of inflation predicted by this class of models would lie between –2 and –4%. This prediction is clearly at odds with observed inflation targets. A second important result that emerges in this class of models is that the Friedman rule is optimal regardless of whether the government is assumed to finance its budget via lump-sum taxes or via distortionary income taxes. This result has been given considerable attention in the literature because it runs against the conventional wisdom that in a second-best world all goods, including money holdings, should be subject to taxation.

One way to induce optimal policy to deviate from the Friedman rule in this type of model is to assume that the tax system is incomplete. We study three sources of tax incompleteness that give rise to optimal inflation rates above the one consistent with the Friedman rule: untaxed profits due to decreasing returns to scale with perfect competition in product markets, untaxed profits due to monopolistic competition in product markets, and untaxed income due to tax evasion. These three cases have in common that the monetary authority finds it optimal to use inflation as an indirect levy on pure rents that would otherwise remain untaxed. We evaluate these three avenues for rationalizing optimal deviations from the Friedman rule both analytically and quantitatively. We find that in all three cases the share of untaxed income required to justify an optimal inflation rate of about 2%, which would be in line with observed inflation targets, is unreasonably large (above 30%). We conclude that tax incompleteness is an unlikely candidate for explaining the magnitude of actual inflation targets.

Countries whose currency is used abroad may have incentives to deviate from the Friedman rule as a way to collect resources from foreign residents. This rationale for a positive inflation target is potentially important for the United States, the bulk of whose currency circulates abroad. Motivated by these observations, we characterize the optimal rate of inflation in an economy with a foreign demand for its currency in the context of a model in which, in the absence of such foreign demand, the Friedman rule would be optimal. We show analytically that once a foreign demand for domestic currency is taken into account, the Friedman rule ceases to be Ramsey optimal. Calibrated versions of the model that match the range of empirical estimates of the size of foreign demand for U.S. currency deliver Ramsey optimal rates of inflation between 2 and 10% per annum. The fact that developed countries whose currency is hardly demanded abroad, such as Canada, New Zealand, and Australia, set inflation targets similar to those that have been estimated for the United States, suggests that although the United States does have incentives to tax foreign dollar holdings via inflation, it must not be acting on such incentives. The question of why the United States appears to leave this margin unexploited deserves further study.

Overall, our examination of models in which a transactional demand for money is the sole source of nominal friction leads us to conclude that this class of models fails to provide a compelling explanation for the magnitude of observed inflation targets.

The second major source of monetary non-neutrality studied in the literature is given by nominal rigidities in the form of sluggish price adjustment. Models that incorporate this type of friction as the sole source of monetary non-neutrality predict that the optimal rate of inflation is zero. This prediction of the sticky-price model is robust in assuming that nominal prices are partially indexed to past inflation. The reason for the optimality of price stability is that it eliminates the inefficiencies brought about by the presence of price-adjustment costs. Clearly, the sticky-price friction brings the optimal rate of inflation much closer to observed inflation targets than does the money-demand friction. However, the predictions of the sticky-price model for the optimal rate of inflation still fall short of the 2% inflation target prevailing in developed economies and the 3% inflation target prevailing in developing countries.

One might be led to believe that the problem of explaining observed inflation targets is more difficult than the predictions of the sticky-price model suggest. For a realistic model of the monetary transmission mechanism, it must incorporate both major sources of monetary non-neutrality, price stickiness, and a transactional demand for fiat money. Indeed, in such a model the optimal rate of inflation falls in between the one called for by the money demand friction — deflation at the real rate of interest — and the one called for by the sticky-price friction — zero inflation. The intuition behind this result is straightforward. The benevolent government faces a trade-off between minimizing price adjustment costs and minimizing the opportunity cost of holding money. Quantitative analysis of this trade-off, however, suggests that under plausible model parameterizations, it is resolved in favor of price stability.

The theoretical arguments considered thus far leave the predicted optimal inflation target at least two percentage points below its empirical counterpart. We therefore consider three additional arguments that have been proposed as possible explanations of this gap: the zero bound on nominal interest rates, downward nominal rigidities in factor prices, and a quality bias in the measurement of inflation.

It is often argued in policy circles that at zero or negative rates of inflation the risk of hitting the zero lower bound on nominal interest rates would severely restrict the central bank’s ability to conduct successful stabilization policy. The validity of this argument depends critically on the predicted volatility of the nominal interest rate under the optimal monetary policy regime. To investigate the plausibility of this explanation of positive inflation targets, we characterize optimal monetary policy in the context of a medium-scale macroeconomic model estimated to fit business cycles in post-war United States. We find that under the optimal monetary policy the inflation rate has a mean of –0.4%. More important, the optimal nominal interest rate has a mean of 4.4% and a standard deviation of 0.9%. This finding implies that hitting the zero bound would require a decline in the equilibrium nominal interest rate of more than four standard deviations. We regard such an event as highly unlikely. This statement should not to be misinterpreted as meaning that given an inflation target of –0.4% the economy would face a negligible chance of hitting the zero bound under any monetary policy. The correct interpretation is more narrow; namely that such event would be improbable under the optimal policy regime.

The second additional rationale for targeting positive inflation that we address is the presence of downward nominal rigidities. When nominal prices are downwardly rigid, then any relative price change must be associated with an increase in the nominal price level. It follows that to the extent that over the business cycle variations in relative prices are efficient, a positive rate of inflation, aimed at accommodating such changes may be welfare improving. Perhaps the most prominent example of a downwardly rigid price is the nominal wage. A natural question, therefore, is how much inflation is necessary to grease the wheel of the labor market. The answer appears to be not much. An incipient literature using estimated macroeconomic models with downwardly rigid nominal wages finds optimal rates of inflation below 50 basis points.

The final argument for setting inflation targets significantly above zero that we consider is the well-known fact that due to unmeasured quality improvements in consumption goods the consumer price index overstates the true rate of inflation. For example, in the United States a Senate appointed commission of prominent academic economists established that in the year 1995–1996 the quality bias in CPI inflation was about 0.6% per year. We therefore analyze whether the central bank should adjust its inflation target to account for the systematic upward bias in measured inflation. We show that the answer to this question depends crucially on what prices are assumed to be sticky. Specifically, if non-quality-adjusted prices are sticky, then the inflation target should not be corrected. If, on the other hand, quality-adjusted (or hedonic) prices are sticky, then the inflation target should be raised by the magnitude of the bias. Ultimately, it is an empirical question whether nonquality adjusted or hedonic prices are more sticky. This question is yet to be addressed by the empirical literature on price rigidities.

Throughout this chapter, we refer to the optimal rate of inflation as the one that maximizes the welfare of the representative consumer. We limit attention to Ramsey optimality; that is, the government is assumed to be able to commit to its policy announcements. Finally, in all of the models considered, households and firms are assumed to be optimizing agents with rational expectations.

2 MONEY DEMAND AND THE OPTIMAL RATE OF INFLATION

When the central nominal friction in the economy originates in the need of economic agents to use money to perform transactions, under quite general conditions, optimal monetary policy calls for a zero opportunity cost of holding money. This result is known as the Friedman rule. In fiat money economies in which assets used for transactions purposes do not earn interest, the opportunity cost of holding money equals the nominal interest rate. Therefore, in the class of models in which the demand for money is the central nominal friction, the optimal monetary policy prescribes that the risk-less nominal interest rate — for example, the return on Federal funds — be set at zero at all times. Because in the long run inflationary expectations are linked to the differential between nominal and real rates of interest, the Friedman rule ultimately leads to deflation at the real rate of interest.

A money demand friction can be motivated in a variety of ways, including a cash-in-advance constraint (Lucas, 1982), money in the utility function (Sidrauski, 1967), a shopping-time technology (Kimbrough, 1986), or a transactions-cost technology (Feenstra, 1986). Regardless of how a demand for money is introduced, the intuition for why the Friedman rule is optimal when the single nominal friction stems from the demand for money is straightforward: real money balances provide valuable transaction services to households and firms. At the same time, the cost of printing money is negligible. Therefore, it is efficient to set the opportunity cost of holding money, given by the nominal interest rate, as low as possible. A further reason why the Friedman rule is optimal is that a positive interest rate can distort the efficient allocation of resources. For instance, in the cash-in-advance model with cash and credit goods, a positive interest rate distorts the allocation of private spending across these two types of goods. In models in which money ameliorates transaction costs or decreases shopping time, a positive interest rate introduces a wedge in the consumption-leisure choice.

To illustrate the optimality of the Friedman rule, consider augmenting a neoclassical model with a transaction cost that is decreasing in real money holdings and increasing in consumption spending. Specifically, consider an economy populated by a large number of identical households. Each household has preferences defined over sequences of consumption and leisure and described by the utility function

   (1)

where ct denotes consumption, ht denotes labor effort, and β ∈ (0,1) denotes the subjective discount factor. The single period utility function U is assumed to be increasing in consumption, decreasing in effort, and strictly concave.

A demand for real balances is introduced into the model by assuming that nominal money holdings, denoted Mt, facilitate consumption purchases. Specifically, consumption purchases are subject to a proportional transaction cost s(vt) that is decreasing in the household’s money-to-consumption ratio, or consumption-based money velocity,

   (2)

where Pt denotes the nominal price of the consumption good in period t. The transaction cost function, s(v), satisfies the following assumptions: (a) s(v) is non-negative and twice continuously differentiable; (b) there exists a level of velocity v > 0, to which we refer as the satiation level of money, such that s(v) = s′(v) = 0; (c) (v v)s′(v) > 0 for v v; and (d) 2 s′(v) + vs″;(v) > 0 for all v ≥ v. Assumption (b) ensures that the Friedman rule, that is, a zero nominal interest rate, need not be associated with an infinite demand for money. It also implies that both the transaction cost and the distortion it introduces vanish when the nominal interest rate is zero. Assumption (c) guarantees that in equilibrium money velocity is always greater than or equal to the satiation level. Assumption (d) ensures that the demand for money is a decreasing function of the nominal interest rate.

Households are assumed to have access to one-period nominal bonds, denoted Bt, which carry a gross nominal interest rate of Rt when held from period t to period t + 1. Households supply labor services to competitive labor markets at the real wage rate wt. In addition, households receive profit income in the amount ∏t from the ownership of firms. The flow budget constraint of the household in period t is then given by:

  

(3)

where τt denotes real taxes paid in period t. In addition, it is assumed that the household is subject to the following borrowing limit that prevents it from engaging in Ponzi-type schemes:

   (4)

This restriction states that in the long run the household’s net nominal liabilities must grow at a rate smaller than the nominal interest rate. It rules out, for example, schemes in which households roll over their net debts forever.

The household chooses sequences {cthtvtMtBt}∞t = 0 to maximize Eq. (1) subject to Eqs. (2)–(4), taking as given the sequences {PtτtRtwt, ∏t}∞t = 0 and the initial M− 1 + R− 1B− 1 condition The first-order conditions associated with the household’s maximization problem are Eqs. (2)–(4) holding with equality, and

   (5)

   (6)

  

(7)

where πt = Pt + Pt − 1 denotes the gross rate of price inflation in period t. Optimality condition (5) can be interpreted as a demand for money or liquidity preference function. Given our maintained assumptions about the transactions technology s(vt), the implied money demand function is decreasing in the gross nominal interest rate Rt. Further, our assumptions imply that as the interest rate vanishes, or Rt approaches unity, the demand for money reaches a finite maximum level given by Ct/v. At this level of money demand, households are able to perform transactions costlessly, as the transactions cost, s(vt), becomes zero. Optimality condition (6) shows that a level of money velocity above the satiation level v, or, equivalently, an interest rate greater than zero, introduces a wedge, given by 1 + s(vt) + vts′(vt) between the marginal rate of substitution of consumption for leisure and the real wage rate. This wedge induces house-holds to move to an inefficient allocation featuring too much leisure and too little consumption. The wedge is increasing in the nominal interest rate, implying that the larger the nominal interest rate, the more distorted is the consumption-leisure choice. Optimality condition (7) is a Fisher equation stating that the nominal interest rate must be equal to the sum of the expected rate of inflation and the real rate of interest. It is clear from the Fisher equation that intertemporal movements in the nominal interest rate create a distortion in the real interest rate perceived by households.

Final goods are produced by competitive firms using the technology F(ht) that takes labor as the only factor input. The production function F is assumed to be increasing and concave. Firms choose labor input to maximize profits, which are given by

   (8)

The first-order condition associated with the firm’s profit maximization problem gives rise to the following demand for labor

The government prints money, issues nominal, one-period bonds, and levies taxes to finance an exogenous stream of public consumption, denoted gt and interest obligations on the outstanding public debt. Accordingly, the government’s sequential budget constraint is given by

In this section, the government is assumed to follow a fiscal policy where taxes are lump sum and government spending and public debt are zero at all times. In addition, the initial amount of public debt outstanding, B–1, is assumed to be zero. These assumptions imply that the government budget constraint simplifies to

where τtL denotes real lump-sum taxes. According to this expression, the government rebates all seignorage income to households in a lump-sum fashion.

A competitive equilibrium is a set of sequences {ct, ht, vt} satisfying Eq. (5) and

   (9)

   (10)

   (11)

  

(12)

given some monetary policy. Equilibrium condition (9) states that the monetary friction places a wedge between the supply of labor and the demand for labor. Equilibrium condition (10) states that a positive interest rate entails a resource loss in the amount of s(vt)ct. This resource loss is increasing in the interest rate and vanishes only when the nominal interest rate equals zero. Equilibrium condition (11) imposes a zero lower bound on the nominal interest rate. Such a bound is required to prevent the possibility of unbounded arbitrage profits created by taking short positions in nominal bonds and long positions in nominal fiat money, which would result in ill-defined demands for consumption goods by households. Equilibrium condition (12) results from combining the no-Ponzi-game constraint (4) holding with equality with Eqs. (2) and (7).

2.1 Optimality of the friedman rule with lump-sum taxation

We wish to characterize optimal monetary policy under the assumption that the government has the ability to commit to policy announcements. This policy optimality concept is known as Ramsey optimality. In the context of the present model, the Ramsey optimal monetary policy problem consists of choosing the path of the nominal interest rate that is associated with the competitive equilibrium that yields the highest level of welfare to house-holds. Formally, the Ramsey problem consists of choosing sequences Rt, ct, ht, and vt, to maximize the household’s utility function given in Eq. (1) subject to Eqs. (5) and (9)–(12).

As a preliminary step, before addressing the optimality of the Friedman rule, let us consider whether the Friedman rule, that is,

can be supported as a competitive equilibrium outcome. This task involves finding sequences ct, ht, and vt that, together with Rt = 1, satisfy the equilibrium conditions (5) and (9)–(12). Clearly, Eq. (11) is satisfied by the sequence Rt = 1. Equation (5) and the assumptions made about the transactions cost function s(v) imply that when Rt equals unity, money velocity is at the satiation level,

This result implies that when the Friedman rule holds the transactions cost s(vt) vanishes. Then Eqs. (9) and (10) simplify to the two static equations¹:

and

which jointly determine constant equilibrium levels of consumption and hours. Finally, because the levels of velocity, consumption, and hours are constant over time, and because the subjective discount factor is less than unity, the transversality condition (12) is also satisfied. We have therefore established that there exists a competitive equilibrium in which the Friedman rule holds at all times.

Next, we show that this competitive equilibrium is indeed Ramsey optimal. To see this, consider the solution to the social planner’s problem

subject to the feasibility constraint (10), which we repeat here for convenience:

The reason this social planner’s problem is of interest for establishing the optimality of the Friedman rule is that its solution must deliver a level of welfare that is at least as high as the level of welfare associated with the Ramsey optimal allocation. This is because both the social planner’s problem and the Ramsey problem share the objective function (1) and the feasibility constraint (10), but the Ramsey problem is subject to four additional constraints, namely Eqs. (5), (9), (11), and (12). Consider first the social planner’s choice of money velocity, vt. Money velocity enters only in the feasibility constraint but not in the planner’s objective function. Because the transaction cost function s(v) has a global minimum at v, the social planner will set vt = v. At the satiation level of velocity v the transaction cost vanishes, so it follows that the feasibility constraint simplifies to ct = F (ht). The optimal choice of the pair (ct, ht) is then given by the solution to ct = F(ht) and –Uh(ct, ht)/Uc(ct, ht) = F′(ht). But this real allocation is precisely the one associated with the competitive equilibrium in which the Friedman rule holds at all times. We have therefore established that the Friedman rule is Ramsey optimal.

An important consequence of optimal monetary policy in the context of the present model is that prices are expected to decline over time. In effect, by Eq. (7) and taking into account that in the Ramsey equilibrium consumption and leisure are constant over time, expected inflation is given by πt+ 1 = β < 1, for all t > 0. Existing macroeconomic models of the business cycle typically assign a value to the subjective discount factor of around 0.96 per annum. Under this calibration, the present model would imply that the average optimal rate of inflation is –4% per year.

It is important to highlight that the Friedman rule has fiscal consequences and requires coordination between the monetary and fiscal authorities. In effect, an implication of the Friedman rule is that nominal money balances shrink at the same rate as prices. The policy authority finances this continuous shrinkage of the money supply by levying lump-sum taxes on households each period. In the present model, the amount of taxes necessary to cover the seignorage losses created by the Friedman rule is given by τtL = (1/β − 1)(Mt/Pt). ² For instance, under a real interest rate of 4% (1/β – 1) = 0.04, and a level of real balances of 20% of GDP, the required level of taxes would be about 0.8% of GDP. The fiscal authority would have to transfer this amount of resources to the central bank each year in order for the latter to be able to absorb the amount of nominal money balances necessary to keep the money supply at the desired level. Suppose the fiscal authority was unwilling to subsidize the central bank in this fashion. Then the optimal-monetary-policy problem would be like the one discussed thus far, but with the additional constraint that the growth rate of the nominal money supply cannot be negative, Mt ≥ Mt–1. This restriction would force the central bank to deviate from the Friedman rule, potentially in significant ways. For instance, if in the deterministic model discussed thus far one restricts attention to equilibria in which the nominal interest rate is constant and preferences are log-linear in consumption and leisure, then the restricted Ramsey policy would call for price stability, Pt = Pt–1, and a positive interest rate equal to the real rate of interest, Rt = 1/β.

The optimality of negative inflation at a rate close to the real rate of interest is robust to adopting any of the alternative motives for holding money discussed at the beginning of this section. It is also robust to the introduction of uncertainty in various forms, including stochastic variations in total factor productivity, preference shocks, and government spending shocks. However, the desirability of sizable average deflation is at odds with the inflation objective of virtually every central bank. It follows that the money demand friction must not be the main factor shaping policymakers’ views regarding the optimal level of inflation. For this reason, we now turn to analyzing alternative theories of the cost and benefits of price inflation.

3 MONEY DEMAND, FISCAL POLICY AND THE OPTIMAL RATE OF INFLATION

Thus far, we have studied an economy in which the fiscal authority has access to lump-sum taxes. In this section, we drop the assumption of lump-sum taxation and replace it with the, perhaps more realistic, assumption of distortionary income taxation. In this environment, the policymaker potentially faces a trade-off between using regular taxes and printing money to finance public outlays. In a provoking paper, Phelps (1973) suggested that when the government does not have access to lump-sum taxes but only to distortionary tax instruments, then the inflation tax should also be used as part of an optimal taxation scheme. The central result reviewed in this section is that, contrary to Phelps’ conjecture, the optimality of negative inflation is unaltered by the introduction of public spending and distortionary income taxation.

The optimality of the Friedman rule (and thus of negative inflation) in the context of an optimal fiscal and monetary policy problem has been intensively studied. It was derived by Kimbrough (1986), Guidotti and Végh (1993), and Correia and Teles (1996, 1999) in a shopping time economy; by Chari, Christiano, and Kehoe (1991) in a model with a cash-in-advance constraint; by Chari, Christiano, and Kehoe (1996) in a money-in-the-utility function model; and by Schmitt-Grohé and Uribe (2004b) in a model with a consumption-based transactions cost technology like the one considered here.

The setup of this section deviates from the one considered in the previous section in three dimensions: First, the government no longer has access to lump-sum taxes. Instead, we assume that taxes are proportional to labor income. Formally,

where τht denotes the labor income tax rate. With this type of distortionary tax, the labor supply Eq. (6) changes to

   (13)

According to this expression, increases in the labor income tax rate and in velocity distort the labor supply decision of households in the same way, by inducing them to demand more leisure and less consumption.

A second departure from the model presented in the previous section is that government purchases are positive. Specifically, we assume that the government faces an exogenous sequence of public spending {gt}∞t = 0. As a result, the aggregate resource constraint becomes

   (14)

Implicit in this specification is the assumption that the government’s consumption transactions are not subject to a monetary friction like the one imposed on private purchases of goods. Finally, unlike the model in the previous section, we now assume that public debt is not restricted to zero at all times. The government’s sequential budget constraint now takes the form

  

(15)

A competitive equilibrium is a set of sequences {vtcthtMtBtPt}∞t = 0 satisfying Eq. (2); Eq. (4) holding with equality; Eqs. (5), (7), (8), and (11); and Eqs. (13)–(15), given policies {Rtτth}∞t = 0, the exogenous process {gt}∞t = 0, and the initial condition M–1 + R–1B–1.

As in the previous section, our primary goal is to characterize the Ramsey optimal rate of inflation. To this end, we begin by deriving the primal form of the competitive equilibrium. Then we state the Ramsey problem. And finally we characterize optimal fiscal and monetary policy.

3.1 The primal form of the competitive equilibrium

Following a long-standing tradition in public finance, we study optimal policy using the primal-form representation of the competitive equilibrium. Finding the primal form involves the elimination of all prices and tax rates from the equilibrium conditions, so that the resulting reduced form involves only real variables. In our economy, the real variables that appear in the primal form are consumption, hours, and money velocity. The primal form of the equilibrium conditions consists of two equations. One equation is a feasibility constraint, given by the resource constraint (14), which must hold at every date. The other equation is a single, present-value constraint known as the implementability constraint. The implementability constraint guarantees that at the prices and quantities associated with every possible competitive equilibrium, the present discounted value of consolidated government surpluses equals the government’s total initial liabilities.

Formally, sequences {cthtvt}∞t = 0 satisfying the feasibility condition (14), which we repeat here for convenience,

and the implementability constraint

  

(16)

given (R–1 B–1 + M–1) and P0, are the same as those satisfying the set of equilibrium conditions (2); Eq. (4) holding with equality; Eqs. (5), (7), (8), and (11); and Eqs. (13)–(15). In the Appendix at the end of the chapter the proof of this statement is presented in Section 1.

3.2 Optimality of the Friedman rule with distortionary taxation

The Ramsey problem consists of choosing a set of strictly positive sequences {cthtvt}∞t = 0 to maximize the utility function (1) subject to Eqs. (14), (16), vt ≥ v, and vsts′(vt) < 1, given R–1 B–1 + M–1 > 0 and P0. We fix the initial price level arbitrarily to keep the Ramsey planner from engineering a large unexpected initial inflation aimed at reducing the real value of predetermined nominal government liabilities. This assumption is regularly maintained in the literature on optimal monetary and fiscal policy.

We now establish that the Friedman rule is optimal (and hence the optimal rate of inflation is negative) under the assumption that the production technology is linear in hours; that is, F(ht) = Aht, where A > 0 is a parameter. In this case, wage payments exhaust output and firms make zero profits. This is the case typically studied in the related literature (e.g., Chari et al., 1991). With linear production, the implementability constraint (16) becomes independent of money velocity, vt, for all t > 0. Our strategy to characterize optimal monetary policy is to consider first the solution to a less constrained problem that ignores the requirement v²ts′(vt) < 1, and then to verify that the obtained solution indeed satisfies this requirement. Accordingly, letting ψt denote the Lagrange multiplier on the feasibility constraint (14), the first-order condition of the (less constrained) Ramsey problem with respect to vt for any t > 0 is

  

(17)

Recalling that, by our maintained assumptions regarding the form of the transactions cost technology, s′(v) vanishes at v = v, it follows immediately that vt = v solves this optimality condition. The omitted constraint v t² s′(vt) < 1 is also clearly satisfied at vt = v, since s′(v) = 0.

From the liquidity preference function (5), it then follows that Rt = 1 for all dates t > 0.

Finally, because the Ramsey optimality conditions are static and because our economy is deterministic, the Ramsey-optimal sequences of consumption and hours are constant. It then follows from the Fisher equation (7) that the inflation rate πt – 1 is negative and equal to β – 1 for all t > 1.

Taking stock, in this section we set out to study the robustness of the optimality of negative inflation to the introduction of a fiscal motive for inflationary finance. We did so by assuming that the government must finance an exogenous stream of government spending with distortionary taxes. The main result of this section is that, in contrast to Phelps’s conjecture, negative inflation emerges as optimal even in an environment in which the only source of revenue available to the government, other than seignorage revenue, is distortionary income taxation. Remarkably, the optimality of the Friedman rule obtains independently of the financing needs of the government, embodied in the size of government spending, gt, and of initial liabilities of the government, (R–1 B–1 + M–1)/P0.

A key characteristic of the economic environment studied here that is responsible for the finding that an inflation tax is suboptimal is the absence of untaxed income. In the present framework, with linear production and perfect competition, a labor income tax is equivalent to a tax on the entire GDP. The next section shows, by means of three examples, that when income taxation is incomplete in the sense that it fails to apply uniformly to all sources of income, positive inflation may become optimal as a way to partially restore complete taxation.

4 FAILURE OF THE FRIEDMAN RULE DUE TO UNTAXED INCOME: THREE EXAMPLES

When the government is unable to optimally tax all sources of income, positive inflation may be a desirable instrument to tax the part of income that is suboptimally taxed. The reason is that because at some point all types of private income are devoted to consumption, and because inflation acts as a tax on consumption, a positive nominal interest rate represents an indirect way to tax all sources of income. We illustrate this principle by means of three examples. In two examples firms make pure profits. In one case, pure profits emerge because of decreasing returns to scale in production, and in the other case they are the result of imperfect competition in product markets. In both cases, there is incomplete taxation because the government cannot tax profits at the optimal rate. In the third example, untaxed income stems from tax evasion. In this case, a deviation from the Friedman rule emerges as optimal because, unlike regular taxes, the inflation tax cannot be evaded.

4.1 Decreasing returns to scale

In the model analyzed thus far, suppose that the production technology F(h) exhibits decreasing returns to scale, that is, F″(h) < 0. In this case, the first-order condition of the Ramsey problem with respect to vt for any t > 0 is given by

Where

As before, ψt denotes the Lagrange multiplier associated with the feasibility constraint (14), λ > 0 denotes the Lagrange multiplier associated with the implementability constraint (16), and ξt denotes the Lagrange multiplier associated with the constraint v²ts′(vt) < 1. The satiation level of velocity, v, does not represent a solution of this optimality condition. The reason is that at vt = v the variable μt is negative, violating the optimality condition μt > 0. To see this, note that μt is the sum of three terms. The first term, μt, ψtcts′(vt), is zero at vt = v because s′(v) = 0. Similarly, the third term, µtµtξt[2vts′(vt) + v²ts″(vt)], is zero because ξt is zero, as the constraint 1 − v²ts′(vt) does not bind at v. Finally, the second term of

is negative. This is because under decreasing returns to scale F′(ht)htF(ht) is negative, and because under the maintained assumptions regarding the form of the transactions technology s″(v) is strictly positive at v.³